- From: C. M. Sperberg-McQueen <cmsmcq@blackmesatech.com>
- Date: Thu, 10 Mar 2011 12:22:38 -0700
- To: ht@inf.ed.ac.uk (Henry S. Thompson)
- Cc: "C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>, www-xml-schema-comments@w3.org
On Mar 10, 2011, at 11:32 AM, Henry S. Thompson wrote: > > C. M. Sperberg-McQueen writes: > >>> >>> ... >> >> ... include is defined on >> the component level, not the schema document level. >> >> Are you proposing that we change it to operate on the >> schema-document level? >> >> What do we do about processors which use includes to incorporate >> components stored in other forms? > > I am confused, as usual (you're fallable, I'm confused, it's chronic :-) > > I read 4.2.3 [1] as being very much expressed in terms of schema > documents. At the information level, as always, but _not_ at the > component level. For example, Certainly the xsd:include element occurs only in schema documents, not in schemas. What I meant was that the semantics of inclusion are defined not in terms of an operation on either the schema document containing the xsd:include nor on the resource identified by the schemaLocation attribute, but in terms of the schema corresponding to the target of the schemaLocation attribute. I thought that was the reason that the 1.0 spec defined its own inclusion mechanism and rejected the suggestions that XInclude be used instead. It wasn't clear to me on first or second reading how the algorithm you propose preserves that property. But perhaps I was conflating the description of include with that of import: the spec is clear that the schemaLocation of an xsd:include must point to an xsd:schema element in an infoset and not to some other representation of a schema. (If I understand correctly, it is not required that schemaLocation point to an XML document; it is allowed instead to point to some vendor-specific representation of an infoset -- JSON, perhaps, or EXI, or a proprietary binary dump format. That may have led me to overlook the fact that the target of schemaLocation must be a schema document, not a schema, as those terms are defined by the XSD spec.) -- **************************************************************** * C. M. Sperberg-McQueen, Black Mesa Technologies LLC * http://www.blackmesatech.com * http://cmsmcq.com/mib * http://balisage.net ****************************************************************
Received on Thursday, 10 March 2011 19:23:10 UTC